Diabetes often leads to serious complications, including coronary heart disease, kidney disease, and blindness, among others. Previous studies have suggested that low-fat, plant-based diets can have a strongly favorable effect on the management of type 2 diabetes mellitus, as well as on the elevations of body weight and serum cholesterol that often accompany it, reducing the risk of complications, and raising the possibility of reducing or even eliminating medication use for many individuals. Evidence suggests that the dietary recommendations that are most effective in diabetes management may be similar to the low-fat, vegetarian diets that have demonstrated utility in reversing coronary artery blockages. However, no study to date has examined the effect of a low-fat, vegetarian diet as an intervention for diabetes in a substantial number of participants, and most studies using plant-based (near-vegetarian) diets have also included exercise as a major intervention component, making it impossible to separate the effects of physical activity from those of diet or to reach any definitive conclusion as to which type of dietary intervention is best. This study, which follows an encouraging preliminary trial reported in Preventive Medicine in 1999, will test the hypothesis that a low-fat, vegetarian diet yields significant improvements in key indices of diabetic control, including glycosylated hemoglobin, fasting serum glucose and insulin concentrations, microalbuminuria, and medication requirements, as well as in cardiovascular risk factors, such as body weight, serum lipids, and blood pressure, in a 22-week intervention controlled throughout for exercise, with a 1-year follow-up. Sixty-eight volunteers with type 2 diabetes will be randomly assigned to a low-fat, vegan (intervention) diet or a control diet deriving 15-20percent of energy from protein and < 7percent of energy from saturated fats, with carbohydrate and monounsaturated fats together providing 60-70percent of energy intake, based on current American Diabetes Association guidelines. Participants in both groups will be asked to attend weekly meetings for nutrition and cooking instruction and group support, and will be asked not to alter their exercise patterns. Physical activity will be monitored by use of the Bouchard 3-Day Physical Activity Record. (Bouchard 1983) Diets will be assessed at baseline and 11, 22, and 74 weeks, using a 3-day dietary record. Fasting serum glucose will be monitored for the study duration and will be used to adjust medications according to a set protocol. Glycosylated hemoglobin, insulin concentrations, 24-hour urinary albumin, body weight, blood pressure, serum lipids, and related cardiovascular risk factors will be measured at baseline, 22 weeks, and 74 weeks.